[This article belongs to Volume - 55, Issue - 01, 2023]
Gongcheng Kexue Yu Jishu/Advanced Engineering Science
Journal ID : AES-06-02-2023-038

Title : MULTIMODAL BIOMETRIC AUTHENTICATION SYSTEM WITH PRIVACY PROTECTION FOR DATA BASE
Yashavanth T R1, Suresh M2

Abstract :

In biometric systems, securityand privacy present significant difficulties. The confidentiality and integrity of biometric data should be secured from all attacks, especially those that target sensitive data. This paper offers a thorough analysis of several physiological biometric methods. To increase the reliability of authentication, this study examines a multimodal biometric system using a combination of three biometric traits: iris, fingerprint, and face.Initially, a watermarking technology has been employed to improve the security for database. LSB Watermarking technology has been employed to Embed Face, irisand Fingerprint images to improve protection in biometric recognition system. In addition to this author have developed Iris and Face traits using LBP and PCA with SVM. Further, Iris-Face-Finger Print traits have been used to develop a more authenticated multi model System with LBP and PCA separately. In the proposed method privacy protection has been provided for the data base using LSB approach. Our proposed method results with an accuracy of 88% for the combination of LBP and SVM with Iris-Face traits.Similarly, 89% is obtained using PCA and SVM, and 92% using LBP and SVMfor Iris-Fingerprint-Face traits respectively. Security level of the biometric system is measured through the Equal Error Rate and is observed as 0.08 for Iris-Fingerprint-Face traits using LBP.